occasions to teach and supervise them on specific topics that were detected during their caring process. These were also occasions to stimulate and enthuse the staff. However, lack of RNs, a lot of administrative work, and too much time spent on drug distribution led to stress at work (V).
Documentation on care of persons with stroke
Physical improvement was the most common reason given in the patients’ records for choice of placement for those discharged home or to rehabilitation ward (I). Many of the patients discharged to a rehabilitation ward were described as cognitively intact, though this was not often clearly expressed as a reason for further rehabilitation. For nursing home placement, the most common reasons were major stroke and severe functional loss with dense hemiplegia.
The planning for further care among those discharged to nursing homes varied as to long-term rehabilitation. Bowel and bladder incontinence, and eating and feeding problems were seldom recorded in the discharge summary. A copy of a nursing summary was rarely found in the patients’ records (I).
Documentation about handling food on the plate/table, behaviour interpreted as pain as well as descriptions of toilet use were very sparse (IV, V) (Table V). However, localisation of pain was more frequent described in the records as daily notes, for example “pain in the shoulder”..
This was also the case as to interventions against pain, where the notes of drugs against pain dominated. One note expressed the person’s metaforic way of describing the pain. Continued recorded evaluations of pain-relieving treatment were sparse, and for some persons there were none (IV). Two persons with swallowing problems were not recorded. Evaluation of the amount of fluid intake by daily recordings was given about four persons of whom two were tube-fed. Notes on food intake in the nursing records were often vague e.g. “eats poorly”.
Mostly notes on food and eating in nursing records concerned the prescribed diet. No nutritional records or advice from a dietician were found in the records (IV). Described and evaluated interventions appeared seldom in the nursing records and there were few care plans;
mostly daily notes were used (IV, V).
Communicating care through different methods
nursing records. The degree of urinary continence was assessed according to the RAI in 100%
of the patients, compared to 25% in the interviews and nursing records (V). Equally ability to manage toilet use and transfer were assessed in 100% compared to 30% and 65% respectively by interviews and nursing records. Although pain was assessed in 68% of the patients and described in 96% by the interviews and nursing records. The interviews more often described the patients’ resources and gave a less impaired view of the patients, while the nursing records more frequently described patients’ need for help. Moreover, behaviour and body expressions interpreted by the RNs as expressions of pain e.g., anxiety, guarding behaviour, were more frequently described in the interviews (V) as were many of the instances of eating behaviour (IV).
Furthermore, some of the interventions elucidated in a comprehensive way that the RNs made many efforts in their nursing care e.g., positioning, transferring; however, this was almost invisible in the nursing records (III, IV, V). The interviews more frequently expressed a restorative/rehabilitative view of the care given than did the records. Described and evaluated interventions appeared seldom in the nursing records; for example, there were few notes about the patients’ food intake, except for the patients who were tube-fed (III, IV,V).
DISCUSSION
The overall aim of this thesis was to explore health status and healthcare needs among persons with stroke, in relation to care planning at emergency hospitals, as well as health status, nursing care needs and nursing care in those living in nursing homes.
Cognition, mood, eating, feeding, nutrition, pain, elimination and transfer became areas of special focus, showing a high frequency of impairment and disabilities, especially in those discharged to a nursing home or living in one (I-V).
Knowing the person
In the interviews a striking matter was the RNs’ descriptions of their tremendous frustration when they could not understand the person cared for. Such difficulties created a sense of frustration as the RNs believed that this led to difficulties in knowing the person cared for and understanding her/his feelings; and thus they did not know what the person wanted (III).The
studies, where knowing the patient is a process of understanding and treating the person as a unique individual (Radwin, 1996. Tanner et al. (1993) presented five aspects of knowing the patient: responses to therapeutic measures, routines and habits, coping resources, physical capacities and endurance, and body typology and characteristics. Knowledge of patient’s experiences, behaviours, feelings and perceptions were other categories of the concept described by Radwin, (1995).
Deep knowledge about the person cared for is important for the way in which care is
performed according to Sherwood (1997) and this is established between the nurse and person cared for over time. Several RNs in our study also emphasized the need to be more active in direct clinical care and were of the opinion that during the caring process one got to know the person more in depth (V). When one saw how they looked, whether they radiated wellbeing, how they behaved etc, then one’s capacity to assess their resources and needs was stated to be much better. Nurses with long clinical experience of cognitively impaired persons described how they could identify specific pain cues by knowing the person. Clusters of individually defined pain cues in the persons cared for are shown to become the recognition feature for pain in cognitively impaired elderly persons (Parke, 1998). In addition RNs expressed that they were responsible for the care, yet due to shortages of RNs often had to rely on staff with an often poor education (V).
Knowing the patient is supported by continuity of care, spending time with the person cared for, and providing more than physical aspects of care (Luker et. al. 2000). Thus, shortage of resources often lead to little time spent in direct clinical care and lack of continuity, which most certainly jeopardized the RNs’ ability to get to know the person cared for. In addition some RNs expressed that information from earlier caregivers was not satisfactory, and written reports, for example, from the speech therapist, or of assessment of cognitive state were sometimes difficult to understand (III).
There were few descriptions on the persons’ own experiences of their condition but examples provided were expressions of tremendous fear related to swallowing, discomfort and
embarrassment as to urinary incontinence and a metaphorical expression from a person in pain (IV-V). This could be explained by the fact that the persons with stroke were assessed
addition 22% of persons with stroke have aphasia, recognised as an important predictor of depression (Åström et al. 1993). The consequences of aphasia have also been described as extremely depressing and frustrating connected with anger, chaos, isolation, and as being a person nobody knows how to relate with (Parr et al. 1997), as well as loss of one’s sense of self (Brumfitt, 1993).
When Sundin et al. (2002) studied communication between care providers and persons with aphasia, they found that when care providers lacked knowledge about the person’s abilities, losses and experience of the current situation, as well as knowledge of the patient’s history, they felt insecure. In our study RNs also expressed a feeling of insufficiency and uncertainty as to how to approach the person, as they did not know them well enough (III). Studies have also shown that caregivers who feel they are unable to make contact with severely demented persons experience care as meaningless and therefore spending less time with those persons (Ekman et al. 1991). Thus there may be a risk that those persons become very isolated, having no one to relate to.
To achieve understanding and being understood, especially for those severely
communicatively impaired, an inter-subjective presence in the parties’ meeting is needed, where the persons’ experiences and feelings are shared by the carers in a silent dialogue guided by those feelings (Sundin & Jansson, 2003). The interpretation of facial expression (Asplund et al. 1995), also described by the RNs (III, V), or by completing a puzzle by combining attuned affects with observations of behaviour and with knowledge about the person’s previous reactions (Häggström et al. 1998) are also ways of understanding communication in this group. Furthermore, Athlin & Norberg (1987, 1998) found that the problems with feeding severely demented persons decreased, and interactive actions increased, when they were fed by staff who had a good knowledge of the patient and their eating based on a patient-assignment system instead of a task-assignment system.
Discharge from acute care
The interviewed RNs in the nursing homes expressed criticism about the stroke patients’ bad state of health at arrival to the nursing home, when they were discharged from acute care hospitals. Their descriptions included for example stroke patients who were confused,
depressed, having bad contractures or who just had started the oral intake of fluids and were looking very thin (III-IV). This was in line with what Styrborn & Thorslund (1993) found among elderly patients referred to be discharged from acute care hospitals, where one-third were in need of further rehabilitation, another one-third needed further medical attention, a number of them were not in a stable medical state and 16% died before discharge. Also in our study (I) we found that the stroke patients decided to be discharged directly to a nursing home from the acute care hospital were still in need of special nursing care, medical care and
rehabilitation due to swallowing problems, eating and feeding disorders, mood disorders and other physical impairments. Furthermore, some of these patients were not in a stable state:
eight were somnolent, of whom five died before the discharge.
Impaired cognition influencing all mental and bodily functions of the patient (Tatemichi et al.
1994) as well as a state of acute confusion, which is common during the first weeks after the stroke onset (Gustafson et al. 1991), might have made the assessment more uncertain for these patients. Notes about difficulties in reaching agreements between the physicians and the representatives from the nursing home, whether the patients were medically ready for
discharge or not, were also found in patients’ records (I). It was concluded from the study by Styrborn & Thorslund (1993) that it is crucial when and on what grounds the physician judges a patient to be medically ready for discharge from an acute care hospital, whether other sections of the care system were prepared to take over and if those persons needs could be met.
Earlier research emphasized longer time as optimal to ensure a good discharge decision for severely impaired stroke patients (Karla, 1993; Hulter-Åsberg & Nydevik, 1991). Karla &
Crome (1993) concluded that an assessment one week after the onset of a stroke may predict the possibility to be discharged home early but is not predictive enough for patients needing longer rehabilitation. The findings in our study indicate that discharge decisions from acute care hospitals for the most severely impaired stroke patients were made after only a few days, as they had a very short mean length of stay at discharge decision (I).
According to the complexity of disorders and the care needs of many persons with stroke, it seems to be especially important with a discharge decision based on knowledge from different
wishes are considered as well as how they may be met. However, for most of the persons discharged to a nursing home the length of stay at the time of discharge decision was probably too short to give the team members sufficient time to meet and interact with the persons and their relatives as well as discuss the affected persons’ different problems from different professionals’ points of view and plan together for further care. Although these persons had the highest mean age and were most impaired, a geriatric consultant was only used in 33% of the cases (I).
The benefits of stroke units with special stroke team maintaining vital functions, preventing deterioration and early mobilisation is evident and well documented (Stroke Units Trialists’
Collaboration, 2000; Jorgensen et al. 2000). However, there is no evidence that treatment in a stroke unit is beneficial if the care is limited to one week (National Board of Health and Welfare, 2000). Furthermore, according to Jorgensen et al. (2000) those with most severe stroke appeared to benefit most. Thus, the persons with stroke discharged directly to a nursing home after a very short length of stay (I) might not receive the optimal care.
Proposal for further treatment and rehabilitation should be noted in the discharge summary as well as the actual state of health of the patient, according to the Swedish authorities (SOSFS 1996a). Further care was planned for those discharged to home, such as receiving
rehabilitation at a day care unit or in the patient’s home, as well as receiving help from the municipality. Although it is not possible to draw any conclusion about the number receiving rehabilitation and help (I), it is important to include the restoration of psychological and social abilities according to the needs and desires of the patients and their families. In a report from the National Board of Health and Welfare (2004), where 5,000 persons were followed up in their homes two years after their stroke onset, more than 30% did not receive
rehabilitation though they were in need.
Several studies has also shown that persons’ quality of life is worse compared with a general population the first year after a stroke (Strum et al. 2004) especially regarding physical factors and vitality (Bugge et al. 2001). However, quality of life is complex and difficult to measure.
Wyller & Kirkevold (1999) found when interviewing persons 3 years after stroke that the reduced quality of life was not interpreted as a direct consequence of body changes, but as a result of the individual’s interpretation and evaluation of the changes. It was essential whether
the person compared the present situation to life prior to the stroke, or to a post-stroke reference point.
Tyson and Turner (2000) found when investigating the discharge of stroke patients from hospitals and follow-up that the treatment was all geared towards getting the patient out of hospital, rather than a rehabilitation process and focused on discharge as the end. This way of acting seemed to have been present for some of the patients being discharged to a nursing home (I), probably due to strong economic demands on the acute care hospitals. Lundh and William (1997) found several shortcomings when investigating discharge planning of elderly people from hospitals in Sweden and UK, and they concluded that discharge must be seen as a continuum, as part of the patient’s care. Thus it is important to ensure that patients’ needs can be met at all levels of care by professional care; this is necessary and also emphasized by the growing evidence that various therapies can improve mobility, activities of daily living and quality of life long after a stroke (Clarke & Ahren, 1994; Drummond & Walker, 1994).
Complexity in persons with stroke
The health status according to the RAI among the persons with stroke in our studies showed a high prevalence of very severely or severely impaired cognition (70%) at time of discharge decision (I), which warred between 30-13% in the nursing homes (II, IV-V). One explanation of this variation might be the increased qualification criteria for entering nursing homes in the last past years (Bergmark et. al. 2000), increased rates of cognitive dysfunction in nursing homes have also been reported in USA (Fries et al. 2000). However, when including all persons with stroke assessed as having some degree of impaired cognition the prevalence was high in all except those discharged to home (I-II, IV-V), and there were a great variation of the degree of impaired cognition. In addition cognition is a complex function which in various ways influences the individual’s daily life in various ways in most activities. RNs also
expressed uncertainty and difficulties in knowing the state of patients’ cognition (III).
Depression was assessed in 15% according to the DRS used in Paper II, while only different signs of mood disorders (55%) were assessed in the other papers (I, IV-V), thus this was not possible to compare Depression has been reported from 10 to 60% during the first year after a stroke (Gresham et al. 1997; Gordon & Hibbard, 1997). The rather low prevalence of
were able to make oneself understood. Although RNs’ descriptions of mood disorders correspond with those found by Bennet (1996) while interviewing skilled nurses about this topic, and also with the diagnostic criteria for major depressive disorders (American
Psychiatric Association, 1994) they found these difficult to distinguish from the actual brain damage or as a result of the current situation (III).
Persons with stroke were in most need of extensive assistance in many of the activities, e.g.
transfer due to significantly more loss of will-conducted movements (I, II, V). The low percentage of independence as to toilet use, dressing and personal hygiene among those with stroke and dementia is probably explained by the influence of impaired cognition, added by motor dysfunction among those with stroke (II). The dependency as to eating was high in those with stroke (II, III) as well as in those with other diagnosis (II), when set-up help was included. Impaired sitting balance, hand/arm movement, swallowing problems (Table IV), side vision problems, all known to be associated with eating problems (McLaren &
Dickerson, 2000; Jacobsson et al. 1996 ) dominated among those with stroke. Furthermore McLaren & Dickerson, (2000) also found that impaired arm movement was the most
powerful predictor of protein and energy consumption. Yet, in nearly 60% of the persons with stroke and in over half of those with other diagnosis there was an uncertainty whether they had lost in weight (II). The number of different eating disabilities in individual persons with stroke, varied from one to seven (III).
The prevalence of pain in the persons with stroke in nursing homes varied between 43% (V) to 47% (II), which is lower than was reported from other studies of persons living in nursing homes (60- 88%) (Blomqvist, 2002). However, pain has been reported even up to 93% in long-term care of stroke patients (Nydevik et al. 1993), when including physical examination.
Furthermore, when three sources of information were used the RAI, interviews and
documentation the prevalence increased from 43 to 63% (V). Thus it is reasonable to assume that pain was undetected, a risk that has been reported, especially in those who suffer from impaired speech, disturbed cognition and distorted body image (Pickering et al. 2000;
Kaasalainen et al. 1998). Thus this is important to prevent, detect and treat. Some of the RNs described that it was difficult to decide whether persons with cognitive and communicative impairments were in pain or not (V).
The highest prevalence in urinary incontinence was assessed in persons with stroke
discharged to a nursing home (I), followed by persons with dementia and stroke in nursing homes (II, V). This was not surprising since incontinence was shown to be a clinical factor after stroke that was associated with greater disability among those in nursing homes (Bean et al. 2003). Impaired cognition, communicative or mobility deficit have also been reported to be barriers to maintaining continence (Brittain et al. 1998).
The RAI assessments as a multidimensional instrument elucidated an extreme complexity of different functional impairments, disabilities, mood disorders and other symptoms important to care planning. In addition this complexity is increased by how different impairments, disabilities and health problems may interact with each other. For example pain has been connected with reduced ADL, sleep disturbance, participation in social activities
(Ferrell et al. 1990), depression (Parmelee et al. 1991). Depression might lead to poor food intake (Herbert, 1996) as well as hindering rehabilitation.
Communication about care
A primary goal of the record is continuity in care for the patients and facilitation of follow-up of the care process, especially for chronically ill patients with multiple diseases. Thus the record should hold valid and reliable information (SOSFS 1996b). However, though eating and feeding as well as incontinence were frequent problems in the acute care hospital, little attention was paid to that in the patients’ records (I). RNs have to define these and other care aspects in the patients’ case records for the purpose of further care. A copy of a nursing discharge summary was rarely found in the patients’ case records (I).
Ehrenberg & Ehnfors (1999) found, when reviewing 120 patients’ records in nursing home, no evidence of a systematic and comprehensive assessment of the patients’ problems based on established criteria or the use of any instrument. This was also the case in our study (IV-V) when reviewing the nursing records of eating, feeding, nutrition, pain, urinary elimination and transfer, and has also been recognised in nursing homes earlier (Hansebo et al. 1999) and primary care (Törnquist, 1998). Many of the described conditions were not documented at all in the nursing records (IV-V), which could be a risk when staff continuity is bad. A striking result from the study was the very few notes of weight found in the records, especially with
bad nutritional state on arrival. Furthermore there was no adequate documentation of follow-up of the persons’ food or nutritional intake, except for those who were tube-fed.
RNs have been recognised to assess patients more than was recorded (Ehrenberg & Ehnfors, 2001). The authors thought that this had to do with accumulated knowledge about the patient and therefore not noted. In this study (VI-V) some descriptions, for example, behaviour interpreted as pain and nursing care related to the studied domains were also more frequently described in the interviews than recorded. When they were recorded, sometimes a less specific view of the interventions were given compared to interviews (IV-V). Lack of carefully recorded care plans may contribute towards care perhaps not being communicated and thus not recognised. For example the described risk of falling when using the toilet was not recorded, which might have contributed to the number of falls in these persons (V).
Transfer in relation to toilet use was the most common activity related to falls in elderly residents (Jensen et al. 2002). The restorative care, which more often was expressed in interviews than in the records, contributed to making the RNs’ care invisible and accordingly less communicated. There is also a risk of ineffective interventions, were nursing staff may perform care differently, leading to uncertainty in the person cared for and thus hinder progress.
RNs’ care of persons with stroke in a nursing home
When reporting the health status of the person cared for the RNs usually described concrete situations from daily life. Some also described the person’s capability although most of them reported care needs (III-V). Thus for food handling on the table or plate they gave examples from the meal situation where behaviour could be interpreted as related to impaired attention, initiation, conceptualisation, visuospatial ability or memory (IV). Many of these eating disabilities have also been recognised in persons with dementia (Tully et al. 1997) but have not often been described in persons with stroke, probably because there have been few studies.
Depending on the complexity of these persons with stroke there is a need for comprehensive assessments in the studied domains. For example as to eating, Jacobsson et al. (1996, 1997) and Westergren et al. (1999) emphasized the importance of a test meal based on a model
consistencies is observed. Through systematic observations of the person’s eating behaviour, sitting position, types of coughing, RNs receive valuable information for diagnostic reasoning and ideas on suitable interventions. Furthermore detection of different risks is important in order to prevent secondary complications. Dysphagia was a common problem among the studied persons (I, II, IV). Some of the persons were also described as suffering from repeated pneumonia or a great deal of coughing, and were extremely frightened about aspiration their breathing (IV).
Short hospital stays and poor care planning, including insufficient information between healthcare levels, magnify the risk that problems such as dysphagia may not be addressed (I).
Furthermore, many of the studied patients suffered from decreased alertness, attention and impulsiveness as well as fatigue (III) which is reported as risks for aspiration of food and liquids (Fienberg et al. 1990). Thus RNs must become competent in eating assessment including swallowing assessment to avoid aspiration of food and liquids known to be
associated with pneumonia, sepsis, and increased morbidity and mortality (Caruthers, 1990).
In addition eating difficulties with strongest predictive value as to nutritional status were found to be ‘dysphagia’, ‘eats three-quarters or less of served food’, ‘alertness’ and ‘aberrant eating speed’, when stroke patients admitted to a rehabilitation ward were studied according to eating, nutritional status, activities of daily living and pressure ulcers (Westergren et al.
2001); thus indicating that several persons in our study were at risk of undernutrition (IV).
Yet only a few weights were recorded as follow up.
Today there are many guidelines and instruments based on nursing research in most of the domains studied that help RNs to perform comprehensive assessments, including those developed especially for persons cognitively and communicatively impaired. However, in the present study the RNs’ descriptions of the persons’ disabilities and problems were often expressed in ordinary terms, lacking comprehensiveness and a diagnostic reasoning process (III-V). Furthermore there were no descriptions of the use of any guidelines or instruments related to the different areas studied, which probably would have helped them in the
diagnostic process, detection of risks and care planning. This might have been one reason for few expressed causes of the different disabilities, and often unspecific described care given (III-V), although there was a tendency that the few specific descriptions were given by the
There is a growing knowledge about the brain plasticity and the influence of different stimuli.
Thus the poor engagement in social activities shown in our studies is a further threat to these persons’ health and wellbeing (II-III). Persons in nursing homes have also been reported as being occupied by few meaningful activities (Bircall & Waters, 1996; Nolan et al. 1995).
The interviews elucidated that instrumental and expressive care often were combined to meet the persons’ needs (III-IV). The importance of the recognition of both these components in nursing care (Sjöstedt, 1997) and in stroke rehabilitation is growing (Burton, 2003; O’Connor, 2000; Kirekevold, 1997). Recently it was also shown in an experimental study of elderly stroke survivors that when nursing interventions related to affective domains and cognitive domains were added to the usual routine rehabilitation programme in the intervention group, they showed better scores in functional, psychological and emotional variables than those who only received the usual routine programme (Nir et al. 2004). In addition this effect had both short-and long-term effects. In our study the RNs were obstructed by insufficient time spent with the person to get to know her/him well enough. This was also aggravated by the extreme complexity in those persons, often added by the cognitive and communicative impairments. Furthermore, there were expressions about insufficient theoretical and
methodological knowledge among some of the nurses. Thus it was not surprising that RNs’
descriptions of their care interventions often were common and lacked specifics (II-V).
Methodological consideration Interviews
Interviewing RNs about their perception of the person cared for as well as the care given and changes in their health status was a method to get better insight (III-V). As these persons suffer from a variety of impairment and disabilities, our semi-structured questions aimed at getting a holistic view of their health situation, which is the situation that confronts RNs and nursing staff. If the interviews had been about specific topics, the answer might have been more specific. However, asking follow-up questions such as “could you tell me more about this“ or “how do you meet these needs” was a way to richer answers. In an interview situation it might be more special descriptions that stick out and probably the most urgent and for
example more obvious nursing care might not have been addressed. To minimise this the interviewer asked follow-up questions such as “is there any other care needed?”.
In the analysis of the different papers (III-V) texts were abstracted from all interviews, which might risk losing coherence; however by working with all interviews several times during a long period, reading them several times and also listening to them during the analysis, this risk decreased.
In a qualitative study credibility refers to whether results give a credible depiction of the respondents’ reality and not merely the researcher’s subjective view (Polit & Beck 2004). In Papers III-V no confirmation of the results has been obtained from the interviewed. Although when discussing this result with RNs working in nursing homes as well as students practising in nursing homes, they often confirm results from interviews.
Transferability has to do with whether the original context is similar to the context where results are to be transferred. The RNs interviewed represented a wide range of age, and of experience working in elderly care. A few had a specialist competence and they worked at different kinds of the nursing homes. Thus they and the nursing homes may be representative at least for big cites but maybe not for the countryside (III-V).
Dependability refers to the stability of data over time and conditions. Assessing the dependability of the data can be done by carrying out a procedure of stepwise replication.
Another researcher should be able to assume similar results but not the contrary to achieve dependability. In Papers III-V the first author (SK) carried out the categorisation, although the two nurse researchers met regularly during the analysis to discuss the meaning units and the categories. This was an analytical process sometimes included re-sorting and reaching agreement on the categories.
The Resident Assessment Instrument
A multidimensional instrument, the Resident Assessment Instrument, RAI, was used in Papers I-II and IV-V to give a comprehensive description of the persons with stroke as to their health status and care needs. Other instruments, focusing on specific areas, might have